Meta-analysis of shotgun sequencing of gut microbiota in obese children with MASLD or MASH

Alterations in the gut microbiome affect the development and severity of metabolic dysfunction-associated steatotic liver disease (MASLD) or metabolic dysfunction-associated steatohepatitis (MASH). We analyzed microbiomes of obese children with and without MASLD, MASH, and healthy controls. Electronic databases were searched for studies on the gut microbiome in children with obesity with/without MASLD or MASH, providing shotgun-metagenomic-sequencing data. Nine studies and an additionally recruited cohort were included. Fecal microbiomes of children with MASLD (n = 153) and MASH (n = 70) were significantly different in alpha- and beta-diversity (p < 0.001) compared to obese (n = 58) and healthy (n = 132). Species Faecalibacterium_prausnitzii and Prevotella_copri are differentially abundant between obese, MASLD and MASH groups. XGBoost and random forest-models accurately predict MASLD over obesity with an AUROC of 87% and MASH over MASLD with 89%. Pathway-abundance-based models accurately predict MASLD over obesity with an AUROC of 81% and MASH over MASLD with 88%. The composition of the gut microbiome is altered with increasing hepatic fibrosis and concomitant species-abundance increase of Prevotella_copri (p = 0.0082). Machine-learning models discriminate pediatric from adult MASH with an AUROC of 97%. The gut microbial composition is increasingly altered in children with the progression of MASLD toward MASH. This can be utilized as a fecal biomarker and highlights the impact of diet on the gut microbiome for disease intervention.